Anda di halaman 1dari 9

The Australian Economic Review, vol. 34, no. 1, pp.


For the Student

Starting Research
John Creedy*
Department of Economics
The University of Melbourne



This article provides a brief guide for students

undertaking their first piece of research.1 The
activity of research itself and the closely related process of writing a research report or
thesis are so different from the standard work
of students, that it is helpful to set down explicitly some of those things that experienced researchers often take for granted. It might be
useful to refer to this article at regular intervals
during the production of your research paper.
In writing the research report, paper, or thesis the major objective, which cannot be overstated, is the achievement of clarity. You need
to produce a transparent statement of the issues, methods and results. This is in fact much
more difficult than is usually realised. There is
no substitute for a careful study of the writing
styles of exemplary authors. A willingness to
respond to constructive criticisms and suggestions is also essential. You need to develop a
new style of writing which is entirely different
from the one used to write undergraduate essays. This article aims only to provide some
brief practical suggestions for organising the
writing and giving a research paper the appropriate shape or appearance: it must look like
a serious piece of research.2

* This guide arose from a number of lectures given over recent years to Honours and Graduate students in Economics
at the University of Melbourne. I have benefited from comments by Denis OBrien and Sheila Cameron on earlier

The nature of the research process and a brief

description of the main properties of a research
paper are provided in Section 2. Suggestions
for arriving at a research topic and making a
start on research are made in Section 3, which
stresses the importance of the plan. General
features of the research report are described in
Section 4. A research paper must satisfy certain
fundamental scholarly requirements; these proprieties are explained in Section 4, which also
provides some suggestions regarding the basic
layout and appearance of the report and some
comments regarding literature reviews. Some
suggestions regarding the writing process are
given in Section 5, with recommendations regarding features to avoid. Section 6 provides
some checklists, and a brief summary is in Section 7.

The Nature of Research

The major difference between research and

coursework is that it is the responsibility of the
researcher to identify a question; you must
specify the question to be examined and decide
on the approach to be used. The need to say
something new necessarily involves a movement into unknown territory; there is no easy
way to check if the answers are right or if the
right method of attack is being used. Research
therefore involves not only the continual exercise of judgement, but also a degree of confidence. In addition, there is no way to avoid
occasionally following false leads and reaching an impasse, that is, going down dead

The University of Melbourne, Melbourne Institute of Applied Economic and Social Research 2001
Published by Blackwell Publishers Ltd, 108 Cowley Road, Oxford OX4 1JF, UK and
350 Main Street, Malden, MA 02148, USA

Creedy: Starting Research

These aspects combine to ensure that the
work should be more interesting and rewarding
than ordinary coursework, while it also gives
rise to alternating phases of optimism and pessimism. Additionally, it is necessary to convince others that it has been worthwhile. There
are times when all researchers feel overwhelmed by difficulties, and are confused,
anxious and not at all sure that they have anything worth reporting. At other times progress
can seem unusually rapid, often helped by what
can only be described as the substantial role
played by serendipity, the faculty of making
happy discoveries by accident. However, remember that fortune favours the prepared
mind. Experienced researchers simply know
that they will go through these phases. For
those who are carrying out research for the first
time, it is worth anticipating these features and
understanding that their experience is not
Perhaps the most important rule of research
is the following: there is no simple relationship
between inputs of time and outputs of useful results. All research meets difficulties. Overcoming them may take a few minutes or it may take
days or weeks. Successful research requires a
willingness to do whatever is needed in order to
overcome the problem, however unimportant it
may seem at the time. Associated with this rule
is the recognition that everything takes longer
than anticipated.
Research also involves intense concentration
over long periods. It is not possible to return to
a research project casually at irregular intervals
or just when there are no other pressing commitments. It is necessary to allocate regular
times to research and always to keep a project
moving forward. Indeed, concentration has to
be such that it becomes something that is extremely hard to stop thinking about.
A simplistic view of research may be described in terms of a linear model in which the
first stage involves reading as much as possible
on a chosen topic and, after having a brilliant
idea of how to proceed, this is followed by the
analysis. The process is completed by simply
writing up the results. However, progress in research is actually highly non-linear. It involves
a complex process described in terms of a re


peated cycle of writing and returning to further

analyses and reading.
Writing is itself a process of discovery, not
least of the authors own level of understanding. It reveals gaps in the argument and suggests
new avenues of research as well as, importantly, providing an error-trapping process.
Most good research, however narrowly defined it may appear initially, has its own momentum. That is, the process of researching a
particular topic leads to further questions and
issues. The completion of a research paper is
therefore often accompanied by negative feelings that, after all, not much has been achieved.
It is worth remembering that this is simply an
aspect of the general truth that the more we
learn, the more conscious we are of our ignorance. Furthermore, progress is in fact largely
achieved by making a series of small steps,
rather than taking giant leaps.
2.1 The Research Paper
It is not easy to summarise briefly the characteristics of a good research paper. However,
any research paper or thesis must at least satisfy the following requirements.

Demonstrate a clear perception of the research problem, its relation to the bigger
picture and the relevant literature.

(ii) Provide motivation for the research question and the approach used.
(iii) Demonstrate an ability to formulate a useful approach and show good judgement in
selecting techniques and, where relevant,
(iv) Show an appreciation of the value and
limitations of the results.
(v) Indicate the potential for further developments.

Getting Started

Getting started is usually a hard and worrying

part of research for newcomers. You can often

The University of Melbourne, Melbourne Institute of Applied Economic and Social Research


The Australian Economic Review

benefit from discussions with other people.

However, the research topic is your choice, so
do something that you find interesting. Do not
worry about what other students or friends
think of your topic, though you will want to
take advice from experienced researchers who
are familiar with the area and its potential problems. The fact that you set the agenda is an important positive feature of research, but at the
same time it presents an unfamiliar challenge.
3.1 Finding a Research Topic
With the first research exercise, your major
concern is initially the question of how to settle
on a topic. It is important to begin with a welldefined question that is not too broad in scope.
Think in terms of taking a number of small
The following suggestions are designed to
make your process of arriving at a precise topic
reasonably systematic. Remember, however,
that this process may take a long time.

Consider an area of economics that, from

previous studies, you find interesting.
This may, for example, be labour economics, monetary economics, public
finance, or international economics. Identify the types of issue that attract you

(ii) The next stage takes place in a library. Go

to the journals section of the library and
look through the contents pages of key
journals in the chosen area. From there,
identify papers to look at. You should
also look at the general journals. There is
no substitute for getting your hands dirty
in this way. At this early stage, do not
simply undertake computer searches
based on keywords.
(iii) When reading journal articles or other research papers, keep the following points
in mind.
Journal papers are usually terse. They
represent work which has matured over
several years. Hence, a full understand

March 2001

ing of the methods and the significance

of the results can only be obtained after
detailed and extensive study. This involves re-reading them several times.
Investigate whether an earlier version,
in the form of a departmental Discussion Paper, is available. This can often
provide more details.
However, a quick initial read will generally be enough to allow you to identify (i) the main question considered by
the author, (ii) the methods of analysis
used, (iii) the data required and (iv) the
nature of the results. These are the four
major features that should receive your
initial attention.
After this preliminary look at particular
papers, you will judge whether they are
of potential interest. You may reject
several papers in this way before finding one that stimulates you to look
closer. If you continue studying the paper, make notes about other literature
cited in it, data used, analytical methods and principal results.
Even at this early stage, keep orderly
notes about the works you consult, including full bibliographical details.
(iv) There is one fundamental ingredient
without which research will never begin.
That ingredient is curiosity. If you have
this, you will never have a problem finding a research topic. When you read papers, always ask yourself questions, such
Can the approach used in a particular
study be applied to other contexts,
countries or time periods?
What assumptions are implicit? Are all
the assumptions sensible? To what extent might the results be sensitive to the
assumptions? How can they be relaxed? Are there any unnecessary assumptions?

The University of Melbourne, Melbourne Institute of Applied Economic and Social Research

Creedy: Starting Research

Is the econometric approach used the
appropriate one? Have all relevant statistical tests been carried out?
Often, precise data relating to the theoretical concepts are not available. Are
the constructed variables the most appropriate for the task?
Are there any implications of the study
which have not been fully drawn out by
the author? Can these be exploited in
your work?

(ii) Attach a time schedule to your plan. Aim

to finish with several weeks to spare. This
will allow you to leave the paper alone for
a while and then give it a final polish after
returning to it refreshed. You will be surprised by how many small but significant
improvements can be made.
(iii) Start writing immediately. As mentioned
above, writing is itself a process of discovery, revealing gaps in the argument
(and your own understanding) and suggesting new lines of enquiry.

(v) In addition to curiosity, you also need a

willingness and the energy to pursue avenues, even if some of these may lead to
a dead end. Furthermore, you need the
imagination and flexibility to overcome
the many inevitable problems along the
road. You also require good judgement to
select the appropriate techniques of
analysis, to decide which aspects can be
safely ignored and which assumptions are
fundamental for the particular context,
and to assess the value of your results at
each stage. The best research reports will
reflect these qualities.


3.2 The Plan of Attack


Research should not be allowed to drift along

in a haphazard way: planning is crucial. You
should have a plan for the big picture as well
as having daily or weekly lists of things to be

At a very early stage, draw up a detailed

table of contents. This may take several
days, as working out the arrangement of
material is often difficult. The table
should contain all your chapter, section
and subsection titles. This plan allows
you to see the sequence at a glance. In
writing, you will not necessarily move
linearly from the first to the last chapter,
and having a clear view of the arrangement will allow you more easily to write
in the most convenient order, while keeping the overall shape in mind.


Basic Features of a Research Paper

Any piece of research, whether a short paper,

report or thesis, is expected to have a number of
basic features. Some general points, concentrating on introductory material, are described
in Subsection 4.1. Subsection 4.2 lists important characteristics which absolutely must be
present. The general layout, structure and appearance are discussed in Subsection 4.3, and
finally the literature review is examined in
Subsection 4.4.
4.1 General Points
It has been mentioned more than once
that you must specify the precise questions that drive your research. The reader
of your paper has to be made familiar
with these questions and the broad structure of your paper or thesis at an early
stage. You must provide the motivation
for the study and the approach. The following advice to theatrical producers, by
W. S. Gilbert, may appear to be rather
vague, but is worth repeating in this context: Tell em what you are going to do;
let em see you doing it; then tell em
what you have done.

(ii) Your introduction needs to let the reader

know, as quickly as possible, three important things. It must answer the questions
what, why and how. Do not digress,
but say what is already known and signal
what is new about your own work. It is

The University of Melbourne, Melbourne Institute of Applied Economic and Social Research


The Australian Economic Review

worth returning to your introduction at
the last stage in the polishing process.
As suggested by Blaise Pascal, the last
thing one knows in constructing a work is
what to put in first. However, remember
that your introduction should be intelligible to someone turning to the topic for the
first time.

(iii) You need to give the reader a clear view

of where each section or chapter is going.
Provide plenty of signposts, which point
the way forward. These can most easily
be added at a later stage, after the first
draft has been completed. Ensure that you
have provided appropriate linkages between various sections. These help to
clarify the logical structure of your thesis.
(iv) Try to form a clear view of your reader. In
particular, you are not writing a textbook,
so a certain amount of knowledge can be
assumed. However, avoid being too allusive. It is perhaps useful, in getting the
level right, to imagine that you are giving
a seminar presentation to your peers.
(v) Remember that the first draft is not the
final draft: it is simply the start of a long
process of revision. It is worth keeping in
mind Samuel Johnsons statement that
what is written without effort is in general read without pleasure. It is remarkable how small changes to crucial
expressions, or minor rearrangements of
material, can improve the clarity substantially.

March 2001

(ii) State when you are summarising other

peoples arguments. This also helps you
to be explicit about precisely how you
have made modifications and original
(iii) Ensure that quotations are accurate and
give exact page references. Do not alter
quotations by, for example, adding emphasis (with italics). Avoid the use of ellipses (), at least if the material omitted
is part of the argument in the quoted material, as distinct from an allusion or reference.
(iv) For all data sources, full details must be
given, including page numbers. It must be
possible for someone to replicate your results with the minimum of effort in obtaining the same data. Keep fully documented
data files in case you are asked to make the
data available to other researchers.
(v) All bibliographical details in your list of
references, arranged alphabetically by author, must be complete and accurate. A
consistent style must be used regarding
capitalisation, italics, initials, and ordering of material. This will require much
more time than you imagine. Several
styles are used, and each publisher and
journal has its own house style. It is important, having settled on a style, to be
consistent. Investigate style requirements
imposed by your department or university
at an early stage.
4.3 Some Mechanics

4.2 Some Proprieties

There are many aspects of writing research reports, such as style and arrangement, which involve choices. However, there are some things
that must be done to satisfy the minimum requirements of scholarship. These are listed

It is important to pay attention to the mechanics of producing a research paper. For example, you must be consistent in the use of titles
and numbering systems. Make decisions regarding the following aspects at an early stage
as it can be very time consuming to make
changes later.



Always acknowledge earlier work. Give

precise sources of the results, diagrams
and equations of other authors.

You are unlikely to need more than three

levels of titles. These are the chapter titles
(numbered 1, 2, ), section titles

The University of Melbourne, Melbourne Institute of Applied Economic and Social Research

Creedy: Starting Research

(numbered 1.1, 1.2, ), and subsection
titles (numbered 1.1.1, 1.1.2, ). Use a
consistent font, capitalisation, spacing
and position (either centred or against the
margin), so that the reader immediately
identifies the status of the title. Keep the
titles succinct but meaningful. The first
sentence after the title should not rely on
that title for its meaning.
(ii) For tables and figures use a decimal numbering system within chapters (for example, Table 1.1, and so on). Give all tables
and figures succinct descriptive titles.
Refer to all tables and figures in the text.
Produce separate lists of tables and figures after the contents page.
(iii) For equations, use a decimal numbering
system within chapters. Number all equations, even if you do not refer to them.
The numbers are useful when other researchers wish to make reference to the
(iv) Use appendices for extensive data descriptions, longer derivations of analytical results, and for subsidiary analytical
or empirical results. Do not use appendices to define notation.
(v) Use footnotes for groups of references to
literature, or qualifications of the main argument. Do not break a sentence with a
footnote flag. Depending on the style requirements, endnotes may be used instead of footnotes. Be sparing in the use
of footnotes; that is, avoid foot-and-note
(vi) In reporting others work, use the past
tense (as in, economist X found that ).
In indicating the contents of your later
chapters or sections, use the present tense
(as in, section X reports estimates of ).
4.4 The Literature Review
The aim of a literature review is to place your
own work clearly within the larger picture.


While research involves a focus on a narrow

range of questions, it is obviously important to
understand how it relates to wider issues. A
brief review of the existing literature can help
to provide some motivation for your analysis.
In addition, it is only possible to establish a
claim to have extended the literature by making
clear the relevant contributions of others. This
can often be achieved relatively quickly, without creating the need for a separate section or
Sometimes it may be necessary to provide an
extensive review of earlier literature in a separate chapter. This presents a difficult challenge
as it calls for quite a mature and confident approach. Ideally, the discussion of the literature
should be organised along analytical or taxonomic lines. This provides clear criteria for deciding whether, and where, an earlier work
needs to be mentioned. Hence:

Start with a clear statement of the broad


(ii) Distinguish alternative possible

proaches, whose features may be:


analytical, involving a range of assumptions and techniques; or

statistical or econometric, associated
with data constraints and estimation
(iii) Refer to earlier contributions in the context of these different approaches. Some
works may therefore be included only as
part of a list while others, judged to be the
most important, may require further discussion.
(iv) Indicate the strengths and weaknesses, in
your judgement, of the various approaches and explain precisely where
your study fits into the taxonomy.
The main thing to avoid is what might be
called the card index method, which consists
of a dull and poorly organised sequence along
the lines of A said this B said that and C
said .

The University of Melbourne, Melbourne Institute of Applied Economic and Social Research


The Australian Economic Review

The Writing Process

The writing process is much more difficult than

is often recognised, though the production of a
clearly written and readable manuscript can be
very satisfying. Subsection 5.1 provides some
general hints regarding writing and revising research work. Subsection 5.2 makes further suggestions, concentrating on things to avoid.
5.1 Initial Hints
The aim of writing is to achieve clarity. This requires great care and a capacity to read your
own work as if it were written by someone else.
The economist Jacob Viner referred to two
basic types of balderdash. The first, simple
balderdash, arises where the author believes
that he or she understands, but cannot make it
intelligible to the reader. The second, compound balderdash, comes in two varieties. In
one variety, neither the author nor the reader
can make any sense of the text, and in the second variety, the reader thinks he or she understands but the author knows it is meaningless.
Your aim is to avoid such balderdash.

March 2001

stopping before reaching that point, you

will find it much easier to pick up the
work the next time and start again, knowing how it needs to proceed.
(iv) Ask a friend to read your draft. Be careful
to select someone you know to be sympathetic and constructive, as anyone can
find negative things to say, however good
the paper. As George Canning, the 19th
century British Prime Minister, pleaded,
save me from the Candid Friend. However, do not try to defend the indefensible.
Do not fall in love with your own writing.
Be willing to respond to suggestions.
(v) When reading through what you have
written, try to produce a succinct summary of each paragraph. This will help to
determine whether a subtitle is needed, or
whether you should change the order of
the material, or whether anything needs to
be added to improve continuity or clarity.
Ask yourself if it is repetitive. If you cannot summarise the paragraph, delete it!
5.2 Further Suggestions


Recognise that your first draft will not be

the last. There is a story of a visitor to an
English stately house asking how the
splendid lawns are produced; the answer
is simply to sow the seed and then weed
and roll it for five hundred years. An
analogy can be drawn with good, clear

(ii) Re-read as you go along. In particular, before turning to a new paragraph, read the
previous one. Before starting a new writing session, re-read the previous work.
This will help to improve continuity.
Regularly check the linkages between
(iii) Stop writing while at a convenient point,
when it is going well. If your writing is
going well, resist the temptation to keep
going until you have reached the end of
the particular section or chapter, or you
have exhausted your current ideas. By

A research paper is not meant to be read aloud

or to entertain the reader. It should be written in
a calm and clear manner so that the emphasis is
always on the issue at hand. Some suggestions,
largely of features to avoid, are listed here.

Avoid colloquial, conversational and

highly personalised expressions.

(ii) Avoid abbreviations (such as &, dont,

and etc.).
(iii) Avoid personal pronouns (I, we, you, me).
(iv) Avoid antiquated, verbose, pedantic and
pompous language.
(v) Do not be allusive.
(vi) Do not annoy the reader by making gratuitous negative remarks about others

The University of Melbourne, Melbourne Institute of Applied Economic and Social Research

Creedy: Starting Research

(vii) Avoid an excessive use of adjectives and
adverbs. When editing your first draft,
look out in particular for very, extremely and highly, which are usually
best deleted. In addition, had can often
be deleted.
(viii) Do not use metaphors, which usually add
colour at the expense of clarity.
(ix) Be gender neutral. This can easily be
achieved without mixing singular and
plural or overusing he or she.


When you are close to producing a polished

draft, it is useful to go through your manuscript
several times, paying special attention to the
overall structure as well as a number of details
of its appearance. Subsection 6.1 provides a list
of structural features to check. Subsection 6.2
lists some of the mechanical aspects that require careful attention. It is usually best to go
through your manuscript separately for each of
the items mentioned in Subsection 6.2, rather
than trying to cover several at one reading.
6.1 The Structure of the Analysis

Is the problem clearly stated?

(ii) Are hypotheses and assumptions explicit?

(iii) Is the relationship to previous work clear?
(iv) Are the limitations acknowledged?
(v) Are the data fully described and their precise sources given?
(vi) Are the conclusions explicitly stated?
6.2 The Basic Appearance of the Paper

Format: Check the prelims, title pages

and contents pages.

(ii) Headings: Check the consistency of style,

fonts, numbering, and spacing.


(iii) Quotations: Check their accuracy and

page references. A large proportion of
quotations are inaccurate!
(iv) Tables: Check titles, abbreviations and
details needed for interpretation and
(v) Equations: Check numbering and crossreferences.
(vi) References: Are all cited works included?
Do not include those not cited. Are works
in alphabetical order? Is the style consistent? Are all details given, such as volume
number, page numbers, date and place of
publication and publisher?


There is no easy formula for producing good

research papers. You need curiosity, energy,
imagination and flexibility to overcome the inevitable problems. You also require good
judgement to select the appropriate assumptions and techniques of analysis, and to assess
the value of your results. The best research reports will reflect these qualities.
Many challenges must be overcome and
even researchers with considerable experience
cannot avoid going down dead ends, occasionally writing sentences containing one of the
two types of balderdash described above, or
even forgetting to mention their key findings
and assumptions. All work must be checked as
carefully as possible and all drafts must be edited and polished many times, paying close attention to detail as well as the overall shape and
flow of the argument.
All this takes longer than envisaged. When
planning your work, produce a generous estimate, fully allowing for the fact that everything
takes longerthen double the time and add
some more for good measure. This is not an exaggeration!
In developing a style of writing research
papers, a great deal can be learned by close
study of authors who are particularly clear.
You may begin by imitating a style that you
strongly admire and find attractive, but of

The University of Melbourne, Melbourne Institute of Applied Economic and Social Research


The Australian Economic Review

March 2001

course ultimately you need to find your own

voice. However, be warned that, as the jazz
musician Miles Davis once said, sometimes
you have to play a long time before you learn
to sound like yourself.

of these. Examples include the books by Anderson and Poole (1994) and Taylor (1989). The
short guide to clear writing by Gowers (1948)
is still well worth reading today. The AGPS
style manual (1995) is a useful publication.

October 2000



Anderson, J. and Poole, M. 1994, Thesis and

Assignment Writing, J. Wiley, Brisbane.
Gowers, E. 1948, Plain Words: A Guide to the
Use of English, HMSO, London.
Style Manual for Authors, Editors and Printers
1995, AGPS, Canberra.
Taylor, G. 1989, The Students Writing Guide
for the Arts and Social Sciences, Cambridge
University Press, Cambridge.

1. This is not a scholarly paper, so I have taken

the liberty of not following all the instructions
given here. In particular, sources of brief remarks by famous people are not given.
2. There are numerous books devoted to thesis
writing, and it would be useful to consult some

The University of Melbourne, Melbourne Institute of Applied Economic and Social Research